| T O P I C R E V I E W |
| Ferdish |
Posted - 03/14/2017 : 11:41:26 AM
Origin Ver. and Service Release (Select Help-->About Origin): 8.5.1 SR2
Operating System: Windows 7 Professional x-64
Hello there!
I have a bunch data which I need to fit with an user-defined fitting function I've made in Function Builder.
When fitting my data using my defined function, sometimes the program doesn't calculate the standard errors for fitting parameters. There are just two dashes across the whole Errors column on Parameters tab, despite that after several iterations the program announces the fit has converged, with Chi-Sqrr tolerance value of 1E-9.
Reduced Chi-Sqr usually varies around the order of 10^-7 and Adjacent R-square about 0.98 or higher in case of my data.
What does it mean? Are those standard errors too high (thousands of percent of the original value) or are they totally off the scale or what? |
| 1 L A T E S T R E P L I E S (Newest First) |
| yuki_wu |
Posted - 03/15/2017 : 04:04:56 AM
Hi,
I supposed that this issue is caused by overflowing the limits of the buffer in the process of calculation. Perhaps you can redefine your fitting function, for example:
Your original fitting function is: y = Ax.
When A is a large value and it may tend to cause overflow, so you can rewrite your function as: y = x / B, in which B = 1/A.
If you still have this issue after you redefine your fitting function, please send your data and FDF file to us via tech@originlab.com, so that we can further check what went wrong.
Regards,
Yuki
OriginLab
|
|
|
